Home // ALLDATA 2024, The Tenth International Conference on Big Data, Small Data, Linked Data and Open Data // View article

An Multi-step Prediction Algorithm for Analysis of Gravitational Waves Based on Deep Learning

Authors:
Chunlin Luo
Yuewei Zhang
Jie Zhu

Keywords: Gravitational Wave; Waveform Prediction; Multistep Prediction; Deep Learning

Abstract:
The detection of Gravitational Wave (GW) events necessitates a vast number of precise GW templates. Improving the accuracy and effectiveness of template waveform synthesis remains a significant challenge. This problem is interesting because accurate templates are critical for identifying GW signals from astronomical events such as Binary Black Hole (BBH) mergers. Previous methods have struggled with cumulative errors and limited parameter ranges, affecting template reliability. We propose a Multi-step prediction method based on deep learning to enhance waveform prediction accuracy. Our approach achieves over 99.6% mean template matching accuracy on a test set of 100,000 waveforms and performs well across a broader parameter range. The main conclusion is that our method significantly improves prediction accuracy and efficiency, facilitating better GW event detection.

Pages: 1 to 6

Copyright: Copyright (c) IARIA, 2024

Publication date: May 26, 2024

Published in: conference

ISSN: 2519-8386

ISBN: 978-1-68558-171-8

Location: Barcelona, Spain

Dates: from May 26, 2024 to May 30, 2024